From: Craig A. Berry [mailto:[EMAIL PROTECTED] > At 2:25 PM +0000 3/12/06, Paul Marquess wrote: > >From: John E. Malmberg [mailto:[EMAIL PROTECTED] > > > >> The test scripts: > >> > >> ext/Compress/Zlib/t/05examples.t and > >> ext/Compress/IO/Zlib/t/010examples.t are both failing on OpenVMS > >> apparently because they can not find the file: > >> auto/Compress/Raw/Zlib/autosplit.ix. > >> > >> The only place I can find a like file is in > >> auto/Compress/ZLIB/autosplit.ix. > > > >Hmmm, not sure what is going on there. > > I'm not seeing any Compress::Zlib failures in blead on VMS, so John's > set-up must be different from mine or the build directory needs some > clean-up.
It could be another ripple for the monster C::Zlib patch I posted last week that moved lots of files around and anyone (myself included) that didn't have their rsync setup correctly ended up with a mish-mash of the before and after C::Z directory tree. > > > ext/Compress/IO/Zlib/t/050interop-gzip.t requires either a shell like > >> bash or use of Perl as a shell for the I/O redirection to work. > > > >Is any I/O redirection available on OpenVMS or is it just piping that > >doesn't work? > > > >For example, would this work? > > > > system "gzip -c infile >outfile"; > > Not unless the gzip program had been built with its own redirection > (like Perl has). To get redirection from the shell you have to use > the pipe command: > > $ pipe show time > foo.tmp > $ type foo.tmp > 12-MAR-2006 12:05:04 Ugh! Would it be better then to just disable the tests that run an external gzip on VMS? Paul